ESPEN Guidelines on Parenteral Nutrition: Pancreas

Assessment of the severity of acute pancreatitis (AP), together with the patients nutritional
status is crucial in the decision making process that determines the need for artificial nutrition.
Both should be done on admission and at frequent intervals thereafter. The indication for
nutritional support in AP is actual or anticipated inadequate oral intake for 57 days. This period
may be shorter in those with preexisting malnutrition. Substrate metabolism in severe AP is
similar to that in severe sepsis or trauma. Parenteral amino acids, glucose and lipid infusion do
not affect pancreatic secretion and function. If lipids are administered, serum triglycerides must
be monitored regularly. The use of intravenous lipids as part of parenteral nutrition (PN) is safe
and feasible when hypertriglyceridemia is avoided. PN is indicated only in those patients who
are unable to tolerate targeted requirements by the enteral route. As rates of EN tolerance
increase then volumes of PN should be decreased. When PN is administered, particular attention
should be given to avoid overfeeding. When PN is indicated, a parenteral glutamine
supplementation should be considered. In chronic pancreatitis PN may, on rare occasions, be
indicated in patients with gastric outlet obstruction secondary to duodenal stenosis or those with
complex fistulation, and in occasional malnourished patients prior to surgery.

Preliminary remarks
Most of the principles of nutritional support in acute and chronic pancreatitis were published in
the ESPEN guidelines on enteral nutrition.1 The aim of this chapter is to focus on specific issues
related to the proper use of parenteral nutrition (PN) in pancreatic diseases. Thus, these
guidelines are an extension of the previously published ESPEN guidelines on enteral nutrition

1. Acute pancreatitis
Introduction
The management, treatment and outcome of acute pancreatitis (AP) are determined by disease
severity. Approximately 75% of the patients have mild disease with mortality below 1%.2
Mortality increases up to 20% if the disease progresses to its severe necrotizing form38 and in
the most severe cases mortality can range from 30 to 40%.
Although the impact of the nutritional status on outcome in patients with severe AP has
not been fully elucidated it is probable that severe malnutrition will adversely affect outcomes9
as occurs in other critical diseases.10 Malnutrition is known to occur in 5080% of chronic
alcoholics and alcohol is a major etiological factor in acute pancreatitis.11 Morbid obesity is also
associated with
poorer prognosis.
Assessment of the severity of AP, together with the patients nutritional status is crucial
in the decision making process that determines the need or otherwise for artificial nutrition. Both
should be done on admission and at frequent intervals thereafter.
The indication for nutritional support in AP is an actual or anticipated inadequate oral intake for
57 days. This period may be shorter in those with pre-existing malnutrition. Assessment of
disease severity assists in this decision because identification of a likely severe course (by
whatever assessment criteria used) alerts the clinician to the likely need for nutritional support.
However, it is worthy of emphasis that no severity and prognosis scoring system is infallible13
and the decisions whether to feed or not must be based on the patients progress irrespective of
any classification system.

1.1. Substrate metabolism in acute pancreatitis

Substrate metabolism in severe AP is similar to that in response to severe sepsis or
trauma. There is increased protein catabolism, characterized by an inability of exogenous glucose
to inhibit gluconeogenesis, increased energy expenditure, increased insulin resistance and
increased dependence on fatty acid oxidation to provide energy substrates (A).
Energy needs may differ and change substantially according to severity and stage of the
disease, patients associated diseases, and specific complications occurring during the clinical
course of AP (B).

1.2. Is amino acid infusion needed and safe during AP?

Severe AP is characterized by substantial protein catabolism and increased energy
requirements (B). Parenteral amino acidinfusion does not affect pancreatic secretion or function
(A).
Comments: Severe AP is mainly characterized, from a metabolic stand point, by nitrogen
waste and protein catabolism with subsequently negative nitrogen balance and secondary
malnutrition. 14 Thus, nitrogen supply should be a major objective of nutrition15 even though a
net positive nitrogen balance is difficult to achieve in such severe conditions.16 In fact patients
with AP, similarly to septic subjects, have an impaired capacity for net protein synthesis and are
less sensitive to protein sparing effects of glucose infusion.16 The goal for nitrogen supply
during total parenteral nutrition in severe AP should be 0.20.24 g/kg per day (equivalent to an
amino acid delivery of 1.21.5 g/kg per day). This requirement should be reduced to 0.140.2 g
nitrogen/kg per day in the case of hepatic or renal failure complicating AP. Monitoring urea
excretion may be helpful to tailor the actual nitrogen need.
There are sufficient data to state that intravenous infusion of amino acids does not affect
the pancreatic secretory response.17,18

1.3. Is the composition of the amino acids given with PN relevant to outcome? When PN is
indicated, parenteral glutamine supplementation (>0.30 g/kg AlaGln dipeptide) should be
considered (B).
Comments: Glutamine is the most abundant free amino acid in the body and has a central role in
many metabolic processes (e.g. a vehicle for interorgan transport of nitrogen and carbon
skeletons, a precursor for nucleotides and glutathione, and a regulator of acidbase balance).19
Several studies have documented beneficial effects of AlaGln supplemented parenteral nutrition
in severely ill ICU patients.20,21 In acute pancreatitis, three randomized controlled studies
including a total of 82 patients compared PN with (N 40) and PN without glutamine (N
42).2224 The majority of patients in these studies had moderate pancreatitis and there was an
overall mortality rate of 4.8%. These studies have slight differences in design and aggregation of
the data precluding a meta-analysis. The overall effect on outcome of parenteral glutamine
supplementation was however positive.25 The use of PN with glutamine was associated with a
trend toward a reduction in overall complications (RR 0.68; 95% CI, 0.421.09; p 0.11)
compared with use of PN alone. Two of these studies showed shorter hospital length of stay with
PN with glutamine compared with PN alone. A non-significant reduction of length of stay of 3
days was reported in the Xian-Li et al. study26 (25.3 vs. 28.6 days) and a significant reduction of
4 days in the Ockenga et al. study (21 vs. 25 days).22 Another study, by Sahin et al.,27 reported
that patients receiving TPN enriched with 0.3 g/kg per day of i.v. glutamine had a significant
reduction in complication rate (10 vs. 40% in controls). Clinical and metabolic benefits of early
infusion of parenteral glutamine dipeptide have been recently confirmed by Fuertes-Orozno et
al.28 and by Xue et al.29 in randomized trials.
No data comparing the optimal dose of glutamine in acute pancreatitis are available. As
in other studies in critically ill patientsa dose of >0.20 g/kg per day of L-glutamine (>0.30 g/kg
per day AlaGln dipeptide) should be considered.20
No data are available on parenteral formulas enriched with other types of amino acids
(e.g. BCAA, EAA, arginine), in this particular clinical setting.

1.4. Is carbohydrate infusion needed and safe during AP?

Glucose should be the preferred carbohydrate energy source for several reasons: it is cheap,
ready available and easy to monitor. Moreover its administration is indicated because it may
counteract gluconeogenesis, but meticulous attention is required to avoid hyperglycemia (A). In
this case exogenous insulin is recommended to maintain blood glucose as close as possible to the
normal range (B). Parenteral carbohydrate infusion does not affect pancreatic secretion and
function (A).
Comments: Glucose should represent the preferred energy supply during AP because it can be
easily administered, itmayin part counteract the intrinsic gluconeogenesis from protein
degradation and provides sufficient calories thus avoiding the use of lipids that maybe
contraindicated in specific patients. Nevertheless, in patients with severe AP, as in other
critically ill patients, glucose oxidation reaches the maximal level of 47 mg/kg per min (56
g/kg per day). Exceeding this limit may cause lipogenesis, hypercapnia and
particularlyhyperglycemia. This latter risk is even greater in patients with AP because pancreatic
necrosis and inflammation impair insulin release. Hyperglycemia needs to be corrected by
exogenous insulin administration, keeping the blood glucose level as close as possible to the
normal range30 (B), although there are as yet no specific studies applying this intervention in
acute pancreatitis. Glucose should represent between 50% and 70% of the total calories (C).
Intravenous glucose does not stimulate exocrine pancreatic secretion (A).31,32.

1.5. Is there a role for non-glucose carbohydrates?

There are insufficient data at this time.
Comment: Only one small clinical trial has addressed the issue of non-glucose carbohydrates (a
mixture of fructose and xylitol) in patients with severe AP. Martinez et al.33 compared two PN
solutions containing either glucose or the above carbohydrate mixture. They reported that the use
of this mixture was associated with a significant lower insulin requirement and blood glucose
levels. These data are insufficient to give a recommendation.

1.6. Is lipid infusion needed and safe during AP?

Lipids provide an efficient source of calories. The use of intravenous lipids in pancreatitis is safe
if hypertriglyceridemia is avoided (C). Triglyceride values below 12 mmol/L are recommended
but ideally serum levels should be kept within normal ranges (C). Current best practice
recommendations are to ensure appropriate infusion rates for fat emulsions (from 0.8 to 1.5 g/kg
per day) and temporarily to discontinue infusion if persistent (>72 h) hypertriglyceridemia occurs
(>12 mmol/L) (C).
Comments: Most of the controversies in the field of artificial nutrition in AP are related to the
use of lipids and in particular long-chain triglycerides, because it is still unclear whether
hyperlipidemia is a cause or a consequence of acute pancreatitis or a combination of both.34 The
latter seems more likely, since serum lipids generally normalize spontaneously in a few days
during recovery from acute pancreatitis. Hyperlipidemia in the most severe cases of AP might
reflect greater disturbances of fat metabolism related to SIRS and sepsis rather than the use of
parenteral lipids. Nevertheless, it is of interest that:
Hypertriglyceridemia (HTG) and chylomicronemia, but not hypercholesterolemia, are found in
about 1238% of patients admitted with acute severe pancreatitis.35
In some cohort studies, an association between altered lipid metabolism and pancreatic
injury3638 was hypothesized, but other clinical reports find no correlation between impaired
lipid catabolism/clearance and AP.39,40
Keeping blood triglyceride levels <4.6 mmol/L, in patients with previous HTG-associated AP
can effectively prevent further episodes of pancreatitis.35
In subjects with hyperlipidemia, the course of acute pancreatitis is more often severe.41,42
Mutations in the lipoprotein lipase gene have been identified in patients with HTG-associated
pancreatitis.43,44
Very few reports link, from a pathophysiological point of view, infusion of lipid emulsions
with the onset of AP. Three cases have been reported in which pancreatitis occurred upon
infusion of lipid emulsions, but co-morbidity of these patients and concomitant therapy (e.g.
corticosteroids) preclude any firm conclusions.45
Lipid emulsions do not affect pancreatic secretion.4651 The diagnosis of
hypertriglyceridemia-associated pancreatitis is based on lipemic serum, a serum TG level greater
that 12 mmol/L and the presence of chylomicronemia.5254 The mechanism(s) by which HTG
might lead to pancreatitis is not completely known, but a theory proposed by Havel et al. is
generally favored.55 Hydrolysis of TG in and around the pancreas by pancreatic lipase secreted
by acinar cell leads to accumulation of free fatty acids in high concentrations. Unbound free fatty
acids are toxic and could. produce injury to acinar cells and microvessels. Increased
concentration of lipids in the pancreatic capillaries then causes vessel plugging and leads to
ischemia and acidosis. In the acidotic environment, free fatty acids cause activation of pancreatic
proenzymes, proinflammatory cytokines and free radicals, thus initiating acute pancreatitis.
Other authors consider that chylomicrons may play a more relevant role than triglycerides in the
association between pancreatitis and HTG.53,56 Moreover, circulating lipase and phospholipase
released during acute pancreatitis may cleave triglycerides and raise serum free fatty acids
(FFA). FFA may lead to intravascular sequestration of calcium by creating FFA-albumin
complexes. In fact, hypocalcemia is a frequent finding in patients with acute pancreatitis and
calcium levels below <2 mmol/l are a well-known negative prognostic factor.57,58 The single
most important point in respect of hypertriglyceridemia- associated pancreatitis is that treatment
of the hypertriglyceridemia may dramatically improve outcomes and can prevent further
pancreatic damage.59 It is important to recognize that hypertriglyceridemia in pancreatitis
generally clears within 4872 h when there is no continuing exogenous source of lipids.54 Only
in cases of hypertriglyceridemia-associated pancreatitis should it be necessary to avoid lipid
emulsions if PN is needed.52 The goal is to maintain TG levels within the normal range.53 If the
serum TG level cannot be maintained below 12 mmol/L, drug therapy is indicated to decrease
VLDL production and prevent more severe HTG. Plasma exchange has been used to lower lipid
and pancreatic enzymes levels, and to improve the signs and symptoms of AP.60 Lipoprotein
apheresis seems even more effective because it removes only large molecular weight complexes
from plasma (such as lipoproteins) and retains immunoglobulins, albumin, and clotting factors,
thus reducing the possibility of infection and bleeding.61,62 If lipids are administered, serum
triglycerides must be monitored regularly, but the use of intravenous lipids as part of parenteral
nutrition in severe acute pancreatitis is certainly feasible on condition that hypertriglyceridemia
is avoided.45,63 At present there are no strong data to suggest that infusion of alternative
parenteral lipids such as omega-3 fatty acids, olive oil, medium chain triglycerides, or structured
lipids, have major clinical advantages compared to soybean oil emulsion. Only one randomized
double-blind trial byWang et al. (N 40) showed that subjects receiving fish oil emulsion
compared to soybean oil, had a reduced inflammatory response, need of renal replacement
therapy and improved respiratory function.64 Larger studies are required to confirm these
preliminary data.

1.7. Is micronutrient infusion needed in AP?

As in all critically ill patients, a daily dose of multivitamins and trace elements is recommended
(C). Although patients with severe AP have demonstrable deficits of plasma and tissue levels of
several micronutrients, at present there are insufficient data to support supranormal doses (C).

Comments: Du et al., randomized 84 patients with AP into two groups, receiving either 1 or 10
g/day of vitamin C. They demonstrated that subjects treated with the higher dose of vitamin C
had a higher cure rate, fewer complications and shorter hospitalization than the low dose
group.65 These results were not confirmed by a case-control study carried out in 46 consecutive
patients with severe AP. The authors reported no clinical benefit from using a high dose of a
multivitamin mixture and selenium.66 Negative results on relevant outcome parameters were
also reported from a randomized double-blind controlled trial by Siriwardena et al.,67 who tested
a similar treatment protocol in patients with severe AP.

1.8. When is PN indicated in acute pancreatitis?

In cases of mild disease, patients can be fed orally after a short period of starvation if pain has
ceased (A). Spontaneous recovery with resumption of oral intake generally occurs within 37
days, and therefore, there is no need for special nutritional treatment (neither PN nor EN) unless
such patients are malnourished prior to the initial attack, or when a therapeutic period of
starvation is indicated for a period of longer than 57 days (A). In this case EN should be started
as soon as possible. The indication for PN is simple and uncontroversial. All patients in whom
the clinician decides that some form of nutritional support is indicated should have this
commenced by the enteral route. Only in those patients who are unable to tolerate targeted
requirements is PN indicated. PN, therefore, is required only when the gut has failed or
administration of EN is impossible for other reasons (e.g. prolonged ileus, complex pancreatic
fistulae, abdominal compartment syndrome) (B). As EN tolerance increases, the volume of PN
should be decreased. When tolerated EN is associated with improved outcomes compared to PN
(A). PN does not significantly stimulate pancreatic secretion in humans and there is no adverse
effect of PN on pancreatic function (A).6870
Comments: There is only one randomized controlled trial, conducted by Sax et al. in 54 patients
with mild AP,71 that has compared the effects of early PN (within 24 h of admission) with no
feeding. PN did not affect the number of days to oral intake, total hospital stay or number of
complications of pancreatitis. Moreover, the authors found that patients with AP had a
significantly higher rate of catheter related sepsis compared to other groups treated with PN in
the same hospital (10.5 vs. 1.5%). Thus, in mild pancreatitis artificial nutrition within 57 days
appears to have no positive impact on the course of disease and is therefore not recommended
unless such patients are malnourished prior to the initial attack or starvation is indicated for
period longer than 57 days. Fluid and electrolytes can be administered parenterally by standard
peripheral routes as required by balance. If early oral refeeding is not feasible due to persistent
pain or changes in severity during the clinical course of the disease, enteral tube feeding should
be delivered.1,72 Normally, most of the patients with predicted severe AP can be fed enterally
and reach the target energy requirement within 34 days from admission.73 If it is predictable
that enteral infusion cannot be started early or not fully tolerated, PN should be started as soon as
possible and a combination of PN and EN is advisable (B). The advantages of the enteral route
over the parenteral one in severe AP have been extensively discussed in the ESPEN guidelines
on enteral nutrition.1 Since the publication of these guidelines, other systematic reviews have
been published.7476 They too confirm the superiority of EN on relevant outcome parameters,
particularly in patients who tolerate enteral infusion. The systemic inflammatory response
associated with severe acute pancreatitis increases metabolic demands and may progress to
multiple organ dysfunction syndromes. The disease and its complications are associated with the
release of proinflammatory cytokines, activation of the complement cascade, release of oxygen
derived free radicals and nitric oxide, and generation of prostaglandin E2, thromboxane A2 and
leukotriene-4 from the metabolism of arachidonic acid. These changes, if prolonged and
combined with starvation lead to a rapid loss of lean body mass with associated morbidity and
mortality. Impaired gut barrier function may lead to translocation of bacteria and their products
from the lumen into the circulation.77,78 The effects of catabolism are compounded by an
inability or reluctance to maintain an adequate oral intake, and patients become malnourished
during the course of their illness. Micronutrient and vitamin deficiencies may also be present on
admission or develop during hospitalization. In addition to hypocalcemia, which is seen in up to
25% of patients with severe AP, deficiencies of magnesium, zinc, folate and thiamine have been
described.79 With this background, intuitively, artificial nutrition should improve outcome in
patients with severe acute pancreatitis when compared with those receiving no nutritional
support. Yet, there is only one small randomized controlled study that compared the effects of
enteral nutrition on 13 patients with severe acute pancreatitis with results from 14 patients who
did not receive nutritional support. This study did not show a benefit of nutritional support on
patient outcome, but interpretation is limited by the fact that it was insufficiently powered.80 At
present there is no definitive evidence that artificial nutritional support alters outcome in most
patients with AP unless malnutrition is also a problem. A diagnosis of AP is not therefore in it
self an indication for instituting artificial nutrition. Nevertheless, in severe cases that are
catabolic and/or have and anticipated period of inadequate oral intake (<1000 kcal/day) longer
than 57 days, it is prudent to begin artificial nutrition either via the jejunum or stomach8184 as
soon as possible in order to prevent the clinical consequences of malnutrition. Sometimes, the
establishment and maintenance of jejunal access in patients with severe AP may be problematic,
and it may be difficult to achieve the targeted intrajejunal nutrient delivery within the first few
days, not least because of impaired upper gastrointestinal motility. A combination of EN and PN
is, therefore, a reasonable way to meet metabolic demands in these patients and the amount of
nutrients delivered parenterally can be progressively reduced as larger volumes are tolerated
enterally.85 There are no randomized studies available comparing the effects of EN alone with
EN supplemented with PN. One study86 randomized 100 patients with severe acute pancreatitis
to receive PN alone or individually staged nutritional support (ISNS) which consisted of PN
albumin in stage 1 (9 _ 6 days), PN EN in stage 2 (6 _ 3 days) and EN alone in stage 3 (3 _ 2
days). When compared with the PN regimen, the ISNS regimen significant reduced the incidence
of sepsis or infection from 30 to 8%, intraabdominal infections from 12 to 4%, hepatic
dysfunction from 12 to 4%, days to resuming oral diet from 25 to 18 days, and length of hospital
stay from 30 to 24 days. However, on the whole, the data are insufficient to determine whether
supplemental EN enhances the efficacy of PN in this group of patients. Nevertheless
supplementing PN with 1030 ml/h of an enteral diet perfused into the jejunum or stomach may
help to maintain gastrointestinal integrity and by providing luminal nutrition (C).87
Complications of acute pancreatitis such as persisting ileus,large pseudocysts, intestinal and
pancreatic fistulae, intestinal edema, retroperitoneal edema, pancreatic ascites, pancreatic or
peripancreatic collections and infected necrosis may sometimes make enteral feeding difficult to
conduct and tolerate, and PN should be instituted along with appropriate treatment for these
conditions (C). Several of the above conditions may also predispose to intraabdominal
hypertension or the abdominal compartment syndrome.88,89 Both air and fluid within the
hollow viscera can be responsible. Gastrointestinal ileus is common both among patients at risk
of developing the syndrome and in those who have already developed it.89 A statistical
relationship has been observedbetween maximal intra-abdominal pressure and overall prognosis
in acute pancreatitis.90 Nasogastric and/or rectal drainage, or endoscopic decompression offer
simple and relatively non-invasive methods for reducing intra-abdominal pressure and treating
mild to moderate intra-abdominal hypertension, and in these patients enteral feeding may be
feasible.91 Continued enteral feeding in the presence of increasing bowel edema and ileus may
however worsen the situation. Non-occlusive bowel necrosis can occur in critically ill patients
receiving enteral nutrition92 and this too may be related to worsening of the intra-abdominal
compartment syndrome. PN will often be necessary in patients with acute pancreatitis and intra-
abdominal hypertension who are not tolerating enteral feeding or who are clinically deteriorating
(C).

1.9. Which is the preferred route for PN?

The central route should be preferred to deliver PN when it is needed in pancreatitis (B).
Comments: The American Society of Parenteral and Enteral Nutrition (ASPEN) Guidelines of
200230 recommend a percutaneously inserted catheter advanced into the superior vena cava as
the route of choice for the delivery of PN, with appropriate confirmation of correct location of
the catheter tip.93 Peripheral PN can however be offered if the anticipated period of nutritional
support is less than 14 days, with the advantages of needing less special expertise and lower
morbidity than central venous cannulation. Although this would seem attractive when PN is
needed in AP peripheral PN is prone to provoke thrombophlebitis.94,95 This complication is
normally reduced by the use of lipid-orientated regimens which are less suitable in AP (see
above). The need for more prolonged PN, and the presence of peripheral edema or poor quality
peripheral veins may also make this route unsuitable. As most patients with severe AP will
warrant central venous access for other reasons it will usually be appropriate for PN to be
administered via a dedicated lumen of a central catheter.

1.10. When is PN contraindicated in AP?

In the severe formof AP, if EN is partially or not tolerated, there is no specific contraindication
to start PN as soon as possible (C). PN should be given after an adequate fluid resuscitation and
when the patient has reached a full hemodynamic stabilization (usually 2448 h from admission)
(B). Lipid administration in PN should stopped temporarily if the plasma triglyceride level rises
over 12 mmol/L, and should be avoided completely in cases of HTG-associated AP (C).
Comments: Xian-Li et al.26 randomized 64 patients with severe acute pancreatitis (no definition
of criteria for severity) to receive PN, PN with glutamine or no nutritional support within 2448
h after fluid resuscitation. Patients given PN plus glutamine, and patients given PN alone had
a significantly better outcome than those in the control group when overall complications, length
of hospital stay and mortality were considered. It is not known exactly when nutritional support
was commenced in this study, but the probable delay of at least 1 day, may have led to PN being
started after the peak of SIRS. Therefore, it may be wise to delay the initiation of PN in patients
with severe acute pancreatitis after adequate fluid resuscitation, achievement of complete
hemodynamic stabilization, and the period of the peak inflammatory response has passed
(usually 2448 h from admission)73,96 (B). Such a period is necessary anyway to re-evaluate
prognostic and severity scoring systems and consequently to reassess the metabolic needs of the
patient. There are no specific contraindications to continuing PN in patients with severe acute
pancreatitis once it has commenced except for the lipid infusion component in patients with
persistent hypertriglyceridemia (>12 mmol/L) (B).

1.11. How should weaning from PN be conducted?

Follow general guidelines for PN weaning.
Comments: There are no specific instructions for weaning patients with acute pancreatitis from
PN and the general guidelines for weaning any patient from PN should be followed.
Improvement in the patients condition should prompt the institution of tube feeding or oral
nutritional supplements and there should be a period of overlap. PN should be progressively
weaned and stopped when the patient is able to tolerate the required amount of nutrients by the
enteral or the oral route. Sudden cessation of PN may result in rebound hypoglycemia which can
be prevented by gradual withdrawal. Oral refeeding with a diet rich in carbohydrates and
proteins and low in fats is recommended (C), but no clinical trials on this are available. If the diet
is well tolerated, oral nutrition can be increased progressively.72 Specific oral supplements are
not recommended.

1.12. Which are the potential pitfalls and complications of PN?

The problems are those of PN in general rather than to its use in AP in particular. Particular
attention should be given to avoid overfeeding (B). Patients should receive 25 non-protein
kcal/kg per day increasing to no more than a maximal caloric load of 30 kcal/kg per day. This
limit should be reduced to 1520 nonprotein kcal/kg per day in case with SIRS or MODS and
when a patient is at risk of refeeding syndrome (B).
Comments: Parenteral nutrition is only required in the management of patients with AP when
they are critically ill and have acute intestinal failure precluding EN, or when they have multiple
metabolic problems. Parenteral nutrition in this setting is best provided by a dedicated nutrition
support team and this approach has been shown to reduce the prescription of inappropriate PN
and also to reduce complications.97 If the central route is being used particular attention should
be paid to catheter insertion in order to prevent procedure-related complications. Catheter care
should be meticulous in order to minimize the risk of catheter related sepsis.98 If the peripheral
route is being used cannulae should be changed at regular intervals and prophylaxis against
thrombophlebitis must be given. Patients with acute pancreatitis often have large fluid deficits
and may require aggressive fluid resuscitation. Salt and water overload is common in these
patients and can be aggravated by PN. Meticulous attention to fluid and electrolyte balance is
mandatory. In AP fluid overload may also predispose to abdominal compartment syndrome.89
Patients with severe acute pancreatitis, especially those with a history of chronic alcoholism and
who are malnourished may be at risk of developing the refeeding syndrome.99,100 Particular
attention should be paid to potassium, magnesium, phosphate, thiamine and sodium balance in
these patients. Appropriate supplements should be given to prevent the development of the
syndrome. If the syndrome does develop, it should be recognized early and treatment measures
instituted immediately. Hyperglycemia is common in patients fed parenterally during AP. This
may be the consequence of several factors: insulin resistance, destruction of islet cells, and
excessive carbohydrate support. Tight control of glucose (between 4.4 and 6.1 mmol/L) with
insulin therapy in critically ill patients has been shown to have a benefit on outcome in selected
patient groups (although not specifically in patients with AP) and would therefore appear to
represent good practice.101,102 Yet, aggressive use of exogenous insulin does put the patient at
risk of severe hypoglycemic episodes.103,104 PN carries the risk of overfeeding the patient,
which may adversely affect outcomes.105 Overfeeding is detrimental for cardiopulmonary and
hepatic function and for carbohydrate and lipid metabolism. To avoid overfeeding, PN should be
initiated with a low calorie regimen and built up step by step with progressive increase with a
tight evaluation of patients metabolism. Hyperglycemia, hyperlipidemia and hepatic dysfunction
and steatosis may necessitate temporary discontinuation of PN. Administration of carbohydrates
in this situation might be not appropriate because this aggravates hepatic steatosis, since these
patients have some degree of insulin resistance.79

2. Chronic pancreatitis (CP)

2.1. When is PN indicated in CP?
Malnutrition is frequent in patients with CP due to paininduced anorexia and to continuing
alcohol abuse. Moreover, increased resting energy expenditure is not rare in these subjects. PN
may, on rare occasions, be indicated in patients with gastric outlet obstruction secondary to
duodenal stenosis and in those with complex fistulating disease (C).
Comments: More than 80% of patients suffering CP can be treated adequately with normal food
supplemented by pancreatic enzymes (B) and only about 1015% of all patients will require oral
nutritional supplements. Tube feeding is indicated in approximately 5% of patients with chronic
pancreatitis (C). Therefore, PN is seldom used in patients with CP and its use will generally be
restricted to those with gastric outlet obstruction in whom gastric decompression is necessary
and a tube cannot be introduced into the jejunum. PN may also be indicated in patients with
chronic pancreatitis and complex pancreatic fistula (C). There are no reported series of patients
with chronic pancreatic insufficiency who have been treated with intravenous nutrition for a long
period. PN is mainly performed over the short term, e.g. in apparent severe malnutrition prior to
pancreatic surgery if enteral feeding is not possible.1

2.2. When is PN contraindicated in CP?

No specific contraindications. There are no specific contraindications to PN in CP, but less than
1% of patients with chronic pancreatitis will ever need PN (C).1 Refeeding syndrome may occur
in these malnourished patients if excessive amounts of calories and nitrogen are delivered in the
first days of PN.